Of the over 450,000 patients undergoing instrumented spinal fusion surgery each year in the United States, over 20,000 ultimately require a revision surgery, the majority of which are secondary to insufficient bone quality. Biomechanical problems associated with insufficient bone quality include pseudarthrosis, insufficiency fractures of the proximal adjacent vertebra, subsidence of any intervertebral implant-graft construct, and loosening of an implanted pedicle screw. No objective clinical test is currently available for surgeons to reliably identify a patient's risk of these adverse outcomes. A bone mineral density test is not enough, as normal bone may still have insufficient quality for high stresses near the implant. The overall goal of this Phase-II SBIR project is to clinically validate our virtual stress test (VST) technology for pre-operatively identifying spinal fusion patients who are at high risk of a failed surgery. VST utilizes the CT scan already acquired for the patient?s pre-operative assessment to construct a virtual, three-dimensional model of the fusion construct in order to investigate its mechanical response to loading. Our proposed pre-operative spinal fusion evaluation, based on VST, has three elements: a) Perform VST to assess pedicle screw loosening, endplate subsidence, and vertebral failure, all specific to the patient?s vertebrae and spinal fusion surgical plan. From that information, identify patients at high risk of a failed fusion surgery due to insufficient bone quality; b) For each vertebral level, specify optimal screw sizing and placement for each pedicle, and measure bone mineral density (BMD) and vertebral strength; c) Provide a comprehensive assessment of risk of a future vertebral fracture due to osteoporosis. Having developed the VST technology in our Phase-I study, this Phase-II study seeks to validate its use in the largest ever biomechanics study of spinal fusion patients. This case-cohort study will include 1,500 women and men who underwent spinal fusion surgery at Kaiser Permanente Southern California from 2009?2018, have a pre-surgical CT scan and at least 90 days of follow-up. Cases (n=500) will be patients with an adverse surgical outcome that required re-operation after the index procedure. Half of this cohort will be used to optimize the virtual stress tests for pedicle screw loosening, endplate subsidence, and vertebral failure. Using the remainder of the cohort, VST will be performed blinded to surgical outcomes in order to test the hypothesis that VST predicts failed surgery, independent of BMD and other clinical risk factors. If successful, this study would result in a biomechanics-based clinical tool for pre-operatively evaluating a patient's risk of spinal fusion failure due to insufficient bone quality. Given the options that now exist for addressing such issues, such as use of anabolic bone forming agents or prophylactic bone cement injection, such a tool could lead to a reduction in adverse surgical outcomes and thus fewer re-operations and revisions.